It has been known for past 10 years that the human parvovirus AAV2 integrates into human chromosome 19 (predominantly at a specific site referred to as AAVS1) to generate a latent form of AAV2 infection in human cells. Prior to our studies, three elements were thought to be important to this event, the viral protein Rep, the target site in AAVS1, and the substrate for integration which focused on the viral ITR elements that function as Rep dependent origins of replication. Integration with recombinant vectors based on viral ITR elements is extremely inefficient (less than 1%). We have initiated a series of studies to characterize the biology of efficient AAV integration. We have made several important discoveries, centered on the discovery of a previously unknown integration element (p51EE) that we have identified. Using a relatively simple assay for efficient Rep mediated integration we have found that between 25 and 50% of transduced cells undergo a successful site specific integration event. This proposal is characterizing the products of the integration event, the substrates of the integration system and we are characterizing the molecular mechanisms that mediate this event through genetic strategies. In addition, we are adapting the AAV2 integration system in a manner that will allow it to be used to mediate tissue specific, species specific integration of any desired transgene. Because all data to date indicates the Rep mediated integration event to be free of negative side effects, the system developed in this proposal will not only characterize a very interesting virus host cell latency system, it will also provide an incredibly powerful system for long term stable in vitro and in vivo gene transfer.